This section describes the information that the users of the framework require, and the requirements and the characteristics that should be considered for inclusion in the proposed framework. Finally, the assumptions of the framework are presented.
3.2.1
Preliminary Information
Quantitative evaluation is a technique used to calculate or approximate the size or level of a specific property, and it yields results in numerical form. In the context of business processes, quantitative evaluation includes approaches such as performance analysis and optimisation. As discussed before in Section 2.4, these approaches use different types of metrics such as service levels, throughput times, and resource utilisation. They provide support in determining the presence of weak points and
bottlenecks in the system and the influence of the adoption of an alternative re- sources or different design models. Any specialist who responsible for making a decision regarding any of the above-mentioned issues is a potential user of the pro- posed framework. This might be a business process engineer, a modeller, or even inexpert users.
Different algorithms are applied to models of business processes for performance analysis and optimisation. Applying the algorithm to a specific business process typically requires bespoke, manual effort by the analyst.
3.2.2
Framework requirements
The requirements for a framework used to perform the quantitative evaluation of business processes can be classified from two perspectives: theoretical feasibility, to show its possible suitability in principle, and practical feasibility, to show its automation and usefulness in the real world. These perspective are described in more detail in what follows.
1. Theoretical feasibility: This concerns the ability of the proposed framework to be used in the quantitative evaluation of business processes. The theoret- ical description helps in demonstrating the feasibility of the new framework, in addition to helping identify the aspects that should be automated. The theoretical description of the framework is presented in terms of its design, and from the perspective of a user of this design. This is help the reader to understand the details that are either shown to or hidden from users of the framework.
2. Practical feasibility: The framework has to be used in a way that is most helpful to its users. To address the practical feasibility related to automation of the framework in particular, a design for a software support tools using Java as object oriented programming language (OOPL) that employs the theoretical aspects required to be created. Furthermore, present the framework from the perspective of the designer of the tool who uses this framework will help in implementing a tool that automates the process of quantitative evaluation of business process. Using these tools will minimise the level of interaction with a user which is needed and allows the process to be automated. To achieve
this automation, the tools have to automatically map information from the business process model into the stochastic Petri net model. Then, parameters relevant for performance analysis and optimisation techniques are extracted and then mapped into Matlab so that these techniques can be applied. Giving the above requirements, the proposed framework first needs to be presented theoretically. The theoretical description of the framework is represented by its design from the user perspective and the practical description is represented by consider it from the perspective of a tool designer.
3.2.3
Characteristics of the framework
A new methodology adopted for the design of a generic modelling framework for the quantitative evaluation of business processes faces many challenges, including definition of appropriate modelling techniques and choosing the metrics to be eval- uated. In identifying these metrics, there is then the challenge of presenting their semantics and expressing these in a way that is appropriate to the elements of a stochastic Petri net model. Given these challenges, the following issues have to be considered when designing such a framework:
1. Different types of business process modelling for quantitative evaluation used to achieved by the implementation of the framework.
2. Different types of metrics are used to apply these modelling techniques. 3. A stochastic Petri net model is used to achieve a higher level of modularity. 4. Automation is used to fulfil the function of the framework.
The motivation for considering these characteristics is that the framework must be capable of being used for any type of business process, or in other words it must be generic and it must allow the use of a wide range of metrics for performance analysis and optimisation. Automation is one of the most important characteristics of the framework, and allows the mapping of information about business processes into stochastic Petri nets and extraction of the relevant parameters which is necessary to perform the quantitative evaluation process without the need for the intervention of the user. Hence, the tool takes the business process information as input and maps it into a stochastic Petri net, and then the output of the stochastic Petri net is mapped to Matlab.
3.2.4
Basic assumptions of the framework
The design of the framework is based on assumptions related to the business process model and the stochastic Petri net model. The business process model is supposed to be syntactically valid and pre-defined in machine-readable language. The stochastic Petri net model is assumed to be ready to use and that there a tool exists to represent the model with a standard file format which is readable after the mapping process has been completed. The file is created from the automatic mapping of business process information into the stochastic Petri net. Also, a high-level mathematical language is assumed to be available for use for the solution of the relevant algorithms.